I’m currently prototyping an offline-first dating service called Bind. I share my thought process on ideating an approach to building in a notoriously difficult space.

I’ve become very curious about what the future of online dating will look like. This makes sense given that I’m a single male in my mid-20s who hasn’t had a great time on dating apps (for various reasons). I have a vested interest in understanding the problem & finding an optimal solution.

In algorithm analysis, we constantly ask ourselves “is this the best we can do?”. We get a deep theoretical understanding of the problem, theorize an optimal performance given the constraints (not yet knowing the exact solution), then work towards a concrete solution achieving that performance.

We also aim to identify solutions that cannot exist, or may appear to be possible, but are impossible for one reason or another. Above all, we never settle for a solution that is “good enough”.

In this case, the problem is that of finding a highly compatible longterm partner. The best (online) solution for that today is Hinge (maybe Bumble, not really Tinder). We will try to understand what comes after these services (2025 & beyond), given the condition & desire of users today.

Before we continue (since it is relevant to the credibility of this writing), this is what I look like:

Prior Art

“Broken” vs Endemic

When people say “online dating today is broken,” they often aren’t very clear on “what” is broken. To create a better solution, it’s important we draw a distinction between problems created within online dating services, & problems that arise from the very nature of how humans make mating choices.

The nature of a problem is immutable. If all complaints we can generate reduce simply to human nature & the mating decision process, we have nothing further to act on or investigate.

So, it’s important we can draw a line between dissatisfaction that is endemic to human mating in any context, then zoom in to the issues endemic to individuals making highly selective mating choices in a context of little information (an online profile).

The Dating Marketplace

Human mating operates as a two-sided competitive market (i.e. labor, equity, energy, etc). It is a bit clinical to use the term “market” here, but being precise will be important to us framing the problem correctly.

Every individual has a goal, and a strategy to achieve that goal. Strategy can be consciously known to the individual, or completely subconscious.

A theoretically “perfect” global solution considering every participant’s goals & strategy would not be one where every individual is maximally satisfied. In any competitive two-sided market, there will always be some portion of individuals who, given their strategy:

• completely fail to achieve their goals
• partially fail to achieve their goals
• partially succeed in achieving their goals
• completely succeed in achieving their goals

Any dating service has to decide how “fair” it will be in its matching process, since it must help users move towards their goals. While not necessarily a zero sum game, one user’s gain may lead to another user’s lost opportunity.

We will not attempt to define what “fair” means for an online dating platform, but it is a very important concept to define. Many users’ interests will be at contention with one another. It could mean giving users an even distribution of impressions to potential matches, showing them to matches of relatively equal desirability, clustering their profile impressions to potential matches in geographic regions close to them, etc.

Mate choice, by design, is not fair. Evolution has encoded within us the drive to favor some as mates, and disfavor others.

No matter how fair or unfair the design of a dating service, this core of exclusivity will exist. There will always be a set of individuals left selected out, or only able to attain a partially successful outcome, and subsequently dissatisfied in not meeting their goals.

According to Pew Research, over 50% of online dating users report somewhat negative, to very negative, experiences dating online^[1]. Is this necessarily an actual problem? Is dating broken? Or are online platforms only surfacing data that has never been available to us before about human selectivity.

What should be of most interest to us is the middle of the distribution. Partially satisfied users who could experience better matches, and partially dissatisfied users whose experience can be elevated by better matchmaking. This is where there are things to be done.

In financial markets, different market microstructures result in different distributions of participant success. To move the dating success distribution, you have to fundamentally rethink the user experience that sets the distribution.

The Match Group Machine

Match Group owns ~75% of the online dating market. Depending on source numbers may vary, but it is roughly:

• Tinder (30%)
• Bumble (26%)
• Hinge (18%)
• (then others)
  • Match.com
  • OkCupid
  • Plenty of Fish (POF)
  • Meetic
  • OurTime (for singles over 50)
  • Pairs (popular in Japan and East Asia)
  • Twoo
  • Stir (focused on single parents)
  • BLK (for Black singles)
  • Chispa (for Latinx singles)
  • Lex (a text-based queer dating app)
  • Tantan (popular in China; Match Group holds an investment stake)
  • ... and the list goes on

References

1. Vogels, Emily A., and Colleen McClain. “Key Findings about Online Dating in the U.S.” Pew Research Center, 2 Feb. 2023, www.pewresearch.org/short-reads/2023/02/02/key-findings-about-online-dating-in-the-u-s/.